print pageprinter friendly
Reports: UNI653486-UNI6: Mechanism of Multi-Spectral Infrared Imaging on Water-Oil Mixture

Debing Zeng, PhD, Saint Peter's University

In the past academic year, three undergraduate students with physics major/minor worked on this project. Their initial goal was to gain hands on experiences in the lab by characterizing the oil samples with an optical spectrometer and a thermal infrared (IR) camera that was available immediately in our lab. It was my hope that this preliminary study could inspire them to do scientific research in their further study and career. The team designed the experimental setup for data acquisition, and wrote the necessary codes with a programming language called “Python” to facilitate this study. Below is a summary for their efforts: 1)      Calibration of the IR camera: In order to have accurate temperature measurements through the thermal IR camera, appropriate calibration for the camera is very important. As mentioned in the camera’s user manual, the error percentage of its temperature reading is +/- 2 oC. However when it was tested against thermocouple with tolerance of +/- 0.2 oC, a nice correlation within +/- 1 oC was acquired. During the experiment, a piece of paper was printed out in black by a regular printer and it was then placed on a piece of aluminum plate as the target for the camera. The tip of the thermocouple was attached tightly to the surface of this target. During the temperature data acquisition with the camera, its parameters were changed while the distance of the target away from the camera was fixed at 20 cm with the emissivity of this target set at 1. The data about atmospheric conditions in the lab were entered manually into the camera. The atmospheric data did not seem to create much of a difference in the temperature reading (Fig. 1). However the distance and emissivity had more noticeable effects on the camera’s temperature reading.

Fig 1. Temperature calibration of Infrared Camera with Thermocouple The second attempt to calibrate the IR camera was to use an aluminum plate painted with black ink from a regular ball point pen. This method produced somewhat better results over the above-mentioned calibration method because it provided better thermal contact than the paper did with the plate. Also studied was the relationship between emissivity and temperature measurement by the IR camera. A fourth degree polynomial fit was fit nicely into the data and in the upper range of emissivity above 0.85, where emissivities of most materials lie, there was essentially a linear relationship between the emissivities and measurements by the camera, as shown in Fig. 2.

Fig 2. Temperature response with change in emissivity 2)      Fluorescence of the samples: Three samples of oil, which are Maxiguard Automatic Transmission Fluid, Maxiguard Motor Oil and Texas Crude Oil, were studied with fluorescence. While these samples are physically distinguishable through their colors, they also produced three distinguishable fluorescence spectra (Fig. 3).

Fig. 3 fluorescence spectra from oil samples 3)      Simulation through COMSOL Multiphysics: With a laser pointer with 5mW in power heating on the surface of the oil samples, the temperature response depends upon the thermal capacity and thickness of the oil samples as well as the environmental temperature. A commercially available software called COMSOL Multiphysics was used to simulate this effect. However, the simulation result by COMSOL shows a rise in temperature of about 1K (Fig. 4) due to the laser heating, which is not consistent with the rise of temperature as much as 10K (Fig 5.) from the actual measurements. Further experiments and more accurate simulation are needed to analyze the inconsistency.

Fig 4. Simulation of the temperature increase by 1K in laser induced heating

 

Fig. 5 Rise in temperature by 10K with a laser of 5mW in power